Game A:
In Game A, all classes can contribute reasonably well in all situations. They contribute in slightly different ways, perhaps, but they all contribute. Now sometimes the way they contribute starts getting a bit similar. For example, a fighter and a rogue both contribute to fights by stabbing things in slightly different ways.
Game B: Not all characters contribute equally in all situations. For example, there is a class that is the best at fighting. There is another class that has more non-combat skills than anyone else. There's a class who's whole shtick is that he's got spells that are better than anything anyone else does, but can't be used very often... and when he can't use spells, he's worse than everyone else. That's the system.
Game B is an OBJECTIVELY worse system
Well, I was going to define "objectively" so I could say you may need to reword things, but you went on to say:
I mean "objectively" worse in the literal sense. It isn't a matter of perspective. Games have design goals, and how well they accomplish them can be objectively evaluated.
Okay, so we're weighing it against goals. What is the goal of Game B? Verisimilitude? Good stories? Well, maybe to some (large) degree at times, but I think you hit on it pretty directly:
Games that do that aren't fun for the guy who gets knocked out.
You said it: fun. Fun is probably the ultimate goal. Can you objectively define what is "fun" for me? I mean, not the definition of "fun", but what makes something Fun. Hmm, maybe you can, since you go on to say:
They lead to bad feelings between the players, and miserable evenings.
So, that's not Fun. Okay... what if it doesn't lead to bad feelings between the players, or a miserable evening? What if a group can play with Game B and, despite how
your group feels about it, actually have Fun with it? What if they have more Fun with it than they did with Game A? Is Game A now objectively worse than Game B? Or, perhaps, is a measure of how Fun something is actually subjective, and not objective?
Similarly, if you have a game who's design goals are apparently to let a group of friends collectively roleplay four to five heroic fantasy characters, and relative to the time spent at the table for any given task what you really create is a game in which one heroic fantasy character is awesome, and his four henchmen kind of help out... you've failed. Objectively.
Okay, but what if people that like Game B don't have those design goals? What if a group of players can have Fun being one powerful guy and four not-as-powerful guys? And, more importantly, what makes you think that Fun isn't the primary goal of either Game A or Game B, and that all other goals take a backseat to it? Or, for that matter, that Game B is has the "five balanced heroic characters story" design goal?
D&D needs to be Lord of the Rings, where Aragorn, Gimli, and Legolas all matter. Not Xena, where everybody knows that in terms of screen time doing cool things, Xena > Gabriella > ... .. .. .. .. > Joxter.
What if Lord of the Rings isn't Fun for the people who like Game B (personally, I'd say Lord of the Rings, taken as a whole with all the main characters, is probably closer to Game B, where you have Gandolf and Pippin in the same party [and let's not get sidetracked with the whole GMPC thing])? Again, does this mean that Game A has objectively failed, even if people do have Fun with Game A?
At any rate, again, I like balance. I prefer it, and I'd rather see the game balanced from the beginning, with mods that can upset that, rather than an unbalanced game that you try to balance via mods. That's my preference. But, trying to objectively define Fun (or trying to define the design goals of 5e as creating a heroic fantasy story for five balanced people rather than being Fun) is probably more than a little off. It's not objective, if nothing else. As always, play what you like
